Typically this style pump isn’t intended to be self priming. It requires a steady supply side pressure. This style pump is intended for booster application where it amplifies incoming water pressure. So if you made the supply side hook up to a running water hose you would likely create pressure higher than that of the hose Source: I work on water booster systems for a living.

You need to separate the high-pressure and low-pressure areas. Otherwise, most of the water will just recirculate within each stage.

All the first impellers should be propellers and the las one an impeller. Reasoning: the impeller accelerates the flow outwards thats why the outlet is tangential, the propellers accelerate flow forward thats why the outlet is coaxial to the inlet. Your pump design accelerates every stage towards the outer wall but the inlet of each stage is in the center, the only optimal stage is the last one with the inlet in the center and the tangential outlet. Also consider individual angular velocities for your stages to improve efficiency. Just my two cents I hope this helps, keep up the good work, people like you make complex designs accessible to people who need them :)

I think most multistage pumps have seals between the stages, otherwise you get flow reversal. For a fluid like water you can easily 3d print labyrinth seals. Cool project, I like the idea.

Here's my idea: water leaves the impellers with a lot of angular momentum; The water may have too much centripetal force to flow back inward into the subsequent impellers. If you could straighten the flow (like with guide vanes), you may be able to make the pump MUCH more efficient.

For the silicon part, you might want to apply the silicon on just one side of the half and let it fully cure before proceed to assemble another half. That way, the silicon will act like a custom made gasket and reusable for many times without need to apply new one for each removal. You might want to apply a thin coat of dish washer soap or vegetable oil on another half for easier removal for first time assembly, just in case it become sticky and for ease of removal later on.

Turbine pumps have to be primed. Almost always. Add a boat propeller to the intake. Big, steep angle that churns water straight in, you'll get a vacuum leak without it. Probably. Run the rod straight out the end, add an aggressive impeller if you are pumping garden/grey water.

Great design :) but you should try this test: pump water into a hose, and see how high you can pump the water. That is the true usefulness of a multistage pump!

Failure is the first step to success 💯

Make inlet=that short pipe connection to pump body airtight. Also, check if your pump makes higher pressure by adding some restriction to outlet. Multi stage pumps are for pressure, not flow.

Backflow is high, you need smaller gaps between inlets and outlets to reduce backflow.

I love how you say "water bump"! Keep it up!

I seriously recommend using the sharpie trick on supports, you use support interface (for a better finish) and set Z Gap to 0mm (just like when using soluble supports) then you insert a pause script before the layer is printed on the supports, you paint the supports with sharpie and I recommend brushing the support roof interface layer with watered down white glue so the part sticks well to the supports, plastic really really doesn't want to stick to sharpie, so the supports (even with the thin glue) come off like butter in one whole piece! closest thing to soluble supports you can have for cheap! I discovered this trick recently and oh my God it's opened to many possibilities, I really hate supports, they're not precise and always get stuck, but my method is flawless you should also take a look at how jet engines and turbomolecular pumps work, you want increasingly higher attack angles and I don't think impellers will stack efficciently, you may want another impeller design (like on commercial pumps and turbos) that start out small and flare out at the base for gradually higher and higher speed, they scoop a ton of water at the small end and accelerate it a lot at the big end

Hard work never fail

Офигенный контент многоступенчатого насоса, разберись с потоком воды, почитай о таких насосах, там поток воды надо разворачивать в разную сторону. И зазор минимальный нужен, а если бы заработала то поток воды был бы офигенный, пожалуйста доделаю этот проект.👍

So I've worked a lot with pumps. I'm no pro, but I'm one damn clever cookie and I grew up on a farm that had a well. So I don't have a formal education but I've spent more time working on pumps than I care to admit. Now centrifugal pumps can't operate with air inside them. In short its because the spin applies force on the liquid that is easily dissipated in only one direction... kinda like a tesla valve. Theoretically is can flow in both directions, one just has orders of magnitude less resistance. When you have an incompressible medium, like water, then a spinning impeller will push water from it's center to its edges. That concept is used in centrifugap pumps by having the water inlet be the center of the impeller. Its spinning pushes the water in a general "way" direction. The housing then contains the water in toroidal flow around the impeller, which we also take advantage of by having the outlet in line with and tangential to the impeller. This results in the water thats pushed by the blades pushing on all the surrounding water, and the path of least resistance is the outlet. Your design has the basic shapes and features of a pump nearly perfect, but you didn't consider the roll pressure would play in this set up. To be more precise, you didn't prime your pump which in your case is absolutely necessary. Notice how in my explanation everything pushes on something else, so all the forces are continuous? When air is in the system it doesn't just come out of its own accord. It is WAY less dense than water so it will float to the top of the housing. Then when the water is trying to find a way to dissipate the energy from the impellers shove the path of least resistance is to compress the air in the housing. A few times won't do much, but if you compress an decompress anything ~30 times a second eventually it'll get hot. Hot things can warp, and warped things break. The other issue I saw was that the stages weren't isolated. Each impeller has only one way to impart its force, from the center to the edge. By not restricting the waters ability to flow between each section the impellers create independently spinning rings of water with the force between them balancing out. The only way this set up will work is if each impeller compounds to the next, else they will all be isolated rings of water. Actually I suspect the small volume that actually came out was due to the water balancing between chanbers: if the impellers weren't perfectly spaced then the distribution of forces between any two chambers would be different. This would force the shaft to move back and forth ever so slightly, randomly creating a momentary seal between sections. When that happened a small volume of water would be forced though to the next section in the chain and the seal would be violently broken creating other seals throughout the system. If you ran this continuously the impellers would eventually be bushes along the shaft until they were all balanced, the output decreasing to zero as this happened. Again, I'm not an educated men so this isn't an "educated guess". I'm just a clever cookie, only a simple farm boy tryin his hand at fluid dynamics.

Printing TPU gasket (tiiviste) instead of silicone sealer would make your projects disassemble easier than breaking them apart.

Another way to chain multiple stages is to add a set of stator vanes between each impeller. They should be spiral shaped to turn the spin of the water as it leaves the impeller into a parallel flow with no spin. Also, on the enclosed vane impeller, you want the input and output cross section to be the same. I.e. it needs to be thinner at the edge than the center.

You need to stop the water rotation after each stage outlet, stator vanes, only then can the pressure increase over each stage. I had a similar issue when designing a multistage air pump, where each stage needs to be smaller than the previous due to compression. Water doesnt compress so each step can be similar in size

very nice video about a waterbump

thats a cool water bump

You should 3d print ninjafex gaskets. So it is more easy to assemble and disassemble

Your problem Travis is that there is no cause for the spun water at high radius to slow and loose it's centrifugal effect to pass to the next stage through the exit hole. You need guiding vanes to catch the spinning water and redirect it from circumferential flow to inwards radial flow. Your motor load and your outlet pressure will drastically increase as a result.

He puts metal, he puts actual metal 2:58. loved it.

why you dont use an printed seal with flexible materials? ? and your problem with the first wheels was... you have some air in it... for those pumps you should all air out of the pump housing

To make supports easier to remove, increase the distance between the highest support layer and the part. Usually its 0 or 1 layers, but changing it to 2 layers should make life easier.

Very interesting video! I like that you listen to your community for ideas. I recently started learning about these pumps and from my understanding so far, they are used to boost the static pressure, not the speed or flow, so different impeller sizes or impeller speed-increase for each stage, like some suggested, is not very useful. However, a re-design to seal the impellers or seperate the outlets from the inlets is probably useful, to prevent recirculation and the pump needs to be primed. For testing, maybe you can compare how high the pump can push the water up in a tube against gravity, this might show the increase in static pressure better. Some thoughts about the mechanical design: Maybe design spacers to precisely seperate the impellers on the shaft for easy assembly Some kind of raised rim on the housing to center the two halves (also makes sealing easier? ) Maybe using only two or three stages instead of going all-in on the first test (optional ;)

I can not imagine how this multiple stages would improve the pump over an single stage pump. Never the less the friction will be 5 times higher, while flow rate and pressure are still the same. So its' 5 time less effektive !

Yes you called it at the end. Each chamber had nowhere to go except to dead head and just let water pass through tiny spaces

Is the pump suffering from cavitation? Judging by the foamy water coming out the outlet...

You should fill the volute with water before starting the impeller. This is called "priming". I am sure that with this little tip you pump will work properly. If you want to improve it even more you should differentiate impellers and volutes as the pressure increases.

Great video as always. I really like the fact that you're still improving and listening to advice and critiques. I think a VERY useful part was the diagram of the water flux, very clear. If i may say, i would added the section of the rotor and maybe a compareson whit a short comment (but you could do it at project ended). Still very good job, keep up the good work. Ps the speacking is still improving, you're getting there.

I’m hooked on this dudes accent. I’m curious as to where he’s from.

At the 8 motor Gerbox and green impeller test, isnt it spinning the wrong direction?

Well, inlet and outlet are bypassed in your Design around the impeller. So you are basically blending water, not pumping

In skateboard they would use bearing spacers for that. You don't need to add them, but in your case it would be ideal.

Great video, thank you. You didn’t appear to silicone between each stage of the pump, so you might have water leaking internally, back into the previous stage.

I like your videos asf. Subscribed!

Hi , the multicelular pump can produc a high presure in the outlet but need an initial inlet presur to work , that meen you have to put an tank in high level and connect him with the inlet of this pump , or you have to add an normal pump (volumic pump booster ) befor that multicelular pump .

Research more on how volute casing works and its arrangement for 2nd stage pumps

Your biggest problem is suction in the pump, obviously; but your turning it on without flooding the pump and air can go back into the pump from the port. Just like a syphon, the pump won't work with air in the system.

Sick 5-double turbo m8

I think I can see the problem. There are no gaps around the impellers so the flow is being restricted and giving you a bad pump. Hope this helps.

each stage has it's own bobbling so at the end, u get lots of air instead of liquid.

Add a planetary gear set in between each one so that they all spin at different speeds.

You need flow straighteners immediately after each impeller otherwise centrifugal force of rotating water is forced toward wall of housing and never makes it into the next stage. flow straighteners stop the water rotating after each stage

this pump when running in the bathtub, sounds like a gas turbine locomotive or a really powerful diesel engine or even a possible jet!

Just thinking out loud, shouldnt the water be going into the inlet? These impelers dont look like they would suck the water from the tank, I wouldve imagined water should already be flowing into the inlet to get pumped. P.s. love your videos!

Hi. Thanks a lot. PLEASE, how do you do to avoid water to pass throw the printed PLA? I have made only one part in printed PLA and there were little drops in the walls outside, it's porous ,even at 100% filling.

brother man... i love the way you think... i wish you were my next door neighbor

Seems you know what you are doing and what is happening XD nice job man...

those tests might have been a low flow rate. but what was the pressure?

Good concert, I agree

Idk if you do but let the silicone firm up some for 5-10 mins then torque the screws down. What we do when we rebuild engines at the shop

Awesome video bro. Support from India

Why not try a tangent combination of multiple pumps? Centrifugal force throws it at 90 degrees from first pump. Second pump takes 90 degrees outlet as its inlet and throws at another 90 degrees. And so on. Bit too late for recommendations but probably only you can do this.

Is it me or is it rotating opposite the direction of the volute outlet?

My friend what youbare making is a high pressure pump. I mean you can load a high pressure tank up to 3000 psi.

Please get a lab bench power supply. You can get the for very cheap and then you can create a higher voltage and I think your pump would work at higher rpms.

Maybe if you blend the blades, like a turbo, even you could add a stator disc

I think every stage should have it' own intake and output and then the outputs mounted in parallel and so the intakes

If I may ask, how efficient would this water pump be if used has a turbine generator?

Good test. But it will only work with independent outputs discharging into a manifold.

multiple stages wont improve flow but it does improve pressure.

It seems like it was probably sucking up air with the water so it never achieved full prime. also you might need like a 5hp motor to power that pump without being under severe load.

Each of these stages is indeed a centrifugal pump - accelerating the water toward the outside of the pump. This is why on the final stage it flares out to the outlet. Each stage accelerates the water toward toward the outside and then it has nowhere to go. This is just a single stage pump still. You would have to have some kind of outlet at the outside of each stage redirecting the water toward the center of the next stage. Also as I'm sure you've noticed, centrifugal pumps are not self priming - you need it to be fed by gravity, or start it with water inside.

i strongly recommend prushament, also pla should work better for most of your projects

You could get some flex material and print costume gaskets instead of having to deal with the messy silicone ;)

you should close the gap between each Centrifugal Propeller and housing from entery side, even if it touch the housing you will get best result

I think there is a problem they way you printed the internal volute; I mean the directions that guide the fluid from the "discharge" of the previous impeller the the suction side of the following impeller, which of course would be a little difficult to print😅. But congrats! it is really beautiful pump.😍

Didn't the pump had to be completely full of water? I know it from selling pumps at first you have to fill them with water so the don't want to pull the air. Try it with a one direktion valve on the inlet side

The idea is very cool but I think it hard to make it more efficient than single impeller pump. Water has too little space and too much distance to travel through all the pump body

Try puting an air leaker on the top to get rid of the air that might be the problem.

The flow rate may only be related to the last pump, and multiple pumps directly connected in series will only reduce the speed and waste power.

it doesn't need to be high volume output, its for high pressure, put a pipe on outlet and see how high water will go up

good work great music

"Well this pump is shit." Lol.

Você deve separar os sistemas pressão. No caso, apenas um rotor esta fazendo a devida pressão a vizinhança virou uma perda de carga enorme. Você pode fazer mais um modulo, fazendo um ofset das peças, e criando uma voluta para cada sistema direcionando para cada camara. Alem disto, eu recomendo uma tomada de pressão ao longo do eixo divergente.

The bubbles seem to be a sign that it's not sealed. I'd look at the neck that goes into the water, you sealed everywhere but there, and as it got further into the water, it started working better.

У тебя не правильного расположение сужений на каждой помпе. Чем меньше расстояние между крыльчатка кой и стенкой корпуса тем больше давление. У тебя водозабор начинается с узкого места и заканчивается широким, где падает давление. Для такого типа помп нужно повышенное давление на всасе. Что бы тянуло с ёмкости нужна помпа -уж улитка. Предлагаю разместить первую помпу (улитку) с клапаном рециркуляции по превышению давления. Напор с неё подать на всас твоей центрифужной помпы. Но я бы там поставил крыльчатка по типу обычных турбин

I have an idea for a engine design this might work well for :-D But ive redesigned the impeller blades in a new configuration i think would work really well...

Hey the new supercharger is here!

The water coming out with the green impellers at the end looks _very_ frothy, blades moving way faster than the water itself and chopping/cavitating?

Nice :)

First of all, the single dc motor was rotating in the other direction thats why it didnt even suck a singke drop of water inside

Génial essayer de faire un compresseur de muscle car😎

Great invention! Did you use fresh water?

Проблема в том что поток после первой крыльчатки продолжает двигаться по спирали, поток нужно выпрямлять перед каждой крыльчаткой.

Dans votre box ou vous mettez vos pièces, avec le petit ventilateur, c'est de l acétone? Pour que les vapeurs lissent le plastique en le faisant fondre?

Looks a lot like booster pumps on older firetrucks

Are you sure the shaft is turning in the right direction?

Интересно. Очень интересный проект. Жду продолжения. Только вот почему не получился первый вариант.... А так выглядит продуктивно.

You need a very tight fit between the impellers and the intake side of the casing. Also those impeller are way too large for the power that you have, I think that pump is about a 1,5kw but I can be wrong.

you need to change impeller design open up the top wall so water can flow into the nest stage

I've noticed that if you have the whole impeller or inlet of the impeller in water then it helps suck it up because there's air in it which is allowing me to fling the water so try to submerge maybe the first stage reply to this if you need any more help

You should definitely add some gears ratios, between turbines

Why not use support that dissolves in water?

great

I hope you succeed in part 2

Awesome